Reverse Audio

The reverb presets share a fixed 2.2-second hall decay. Worth knowing: the speed control is varispeed, like slowing down a record, so anything below 1.00x also lowers the pitch. That is fine — often helpful — when you are listening for words, but it means the slowed presets are not pitch-preserving. If you only want a clean backwards copy, stay on Plain Reverse and export as WAV: that path applies no speed change, no filtering, no gain change, and no reverb at all.

What Happens When You Reverse Audio?

A digital audio file is a long list of samples — tens of thousands of amplitude measurements per second. Reversing plays that list from the end to the beginning. Nothing is added, removed, or recalculated, which is why reversing is exact and why reversing a reversed file gives you the original waveform back.

The pitch and tempo do not move — the set of frequencies in the recording is identical either way round. What moves is the envelope of every sound. A piano note normally begins with a hammer strike and decays into silence. Reversed, it begins as near-silence, swells into a crescendo, and cuts off dead at the moment the hammer originally hit. Your ear reads a sharp attack as evidence of a physical cause, so removing it makes each sound seem to arrive from nowhere.

That one change is enough to disguise an instrument completely. In a classic recorded demonstration, a piano performance played back with every note reversed no longer sounds like a piano at all — it sounds like a reed organ. Nothing about the notes changed; only the direction of their envelopes did. It is also why reversed audio pairs so naturally with reverb: a reverb tail is pure decay, so reversing it produces the longest, smoothest swell available without synthesising anything.

Reverse Reverb, Risers, and Build-Ups

The single most useful thing you can do with reversed audio is make a sound arrive before itself. The studio recipe has four steps: reverse the clip, apply reverb to the reversed version, render that to a real audio file, then reverse the result back. The reverb that used to trail after the phrase now swells into it.

Two details decide whether it works. First, you have to bounce the reverb to audio before the second reversal — a live plugin cannot be played backwards, so the tail has to exist as samples first. Second, leave a few seconds of silence after the reversed clip before you add reverb, so the tail has somewhere to decay into; that silence is exactly what becomes the swell's lead-in once you flip it back. A 2–5 second decay covers most sources.

This page handles the reverse-and-reverb half of that chain in one pass: set the Reverb slider, export, and you have the source file to flip back in your editor. For a reverse cymbal riser, let the crash ring until it has fully decayed before you reverse it — a truncated tail produces a riser that starts abruptly instead of emerging from nothing — then place it so it ends precisely where the drop lands. The same trick with a kick drum, timed to finish as the first downbeat kick hits, is a staple of dance transitions. The technique has a long pedigree on record: reverse echo appears on Led Zeppelin's "Whole Lotta Love," Genesis's "Deep in the Motherlode," and the vocal on Metallica's "Fade to Black."

Backmasking vs. Hearing Things: What Reversed Records Actually Contain

Two different things get called "hidden messages," and keeping them apart is most of the story. Backmasking is deliberate: a sound recorded, flipped, and mixed in so it only reads backwards. Phonetic reversal is accidental: ordinary forward audio that happens to resemble words when reversed. Almost every famous case is the second kind.

The deliberate ones are real and easy to verify with this tool. The Beatles' "Rain" (1966) ends on a genuine backwards vocal. Pink Floyd buried a spoken message in "Empty Spaces" on The Wall that congratulates you for finding it. ELO answered their accusers on "Fire on High" with a reversed line about music being reversible while time is not. Missy Elliott made the idea into a hook on "Work It," following "I put my thang down, flip it and reverse it" with the line itself reversed — by her account a studio accident they liked enough to keep.

The notorious cases went the other way. When Judas Priest were sued over an alleged subliminal "do it," the case collapsed in August 1990: the judge found the plaintiffs had not shown the band placed any such message on the record. Guitarist Glenn Tipton had gone back to a studio, played the album backwards himself, and turned up phrases just as convincing and obviously accidental — including "Give me a peppermint" and "Hey ma, my chair's broken." Led Zeppelin's response to the "Stairway to Heaven" claims was shorter: their label pointed out that turntables only play in one direction.

There is research behind why this happens. Reversed speech keeps its speech-like texture while losing its meaning, so it sits in exactly the ambiguous zone where suggestion takes over. In a 1985 study, listeners could not detect backmasked phrases played forwards, could not sort reversed messages by type above chance, and showed no behavioural effect from them — the authors concluded that hearing a backward message is more an act of construction by the listener than a property of the recording. An earlier experiment found that simply telling people what to listen for was enough to make them hear it. So reverse a track and listen for yourself, but expect your own expectations to do most of the work.

The Reverse Singing Challenge — and the People Who Did It Properly

The challenge circulating on TikTok and Reels is simple: reverse a song, learn to sing along with the backwards version, record your attempt, then reverse the recording and hear how close it landed. Reverse the track here, use the 0.75x preset while you are learning the syllables, and export the backwards version to sing against — backwards phonetics at full speed are close to impossible to copy.

The technique long predates the trend, and the professional results are startling. On Twin Peaks in 1990, Michael J. Anderson's dialogue was recorded, reversed, learned by imitation, performed, and reversed again in the edit — the uncanny cadence is phonetic reversal, not an effect. Chris Martin spent a month learning to sing "The Scientist" backwards for its video. The Pharcyde were coached through "Drop" by a linguist who wrote out the reversed phonetics for them, and they had under a week to memorise it. Mutemath rehearsed "Typical" in reverse thoroughly enough to perform the backwards version live on television. All of them were doing exactly what the challenge asks — with more rehearsal.

Reversing vs. Slowing vs. Pitching

Reversing changes direction — the order events happen in — and leaves pitch and tempo alone. Speed changing changes time, and in a simple varispeed tool it drags pitch along with it. Pitch shifting changes key while the timeline stays put. They stack cleanly and in any order: a common edit is to reverse a track here, then run the export through the Slowed Reverb Generator for a backwards, slowed, cavernous version.

Reverse Audio FAQ